veo-3-1/extend-video
Seamlessly lengthen shots with frame-consistent context control and audio blending for refined video creation.
Seedance 2.5 is ByteDance Seed's next-generation multimodal video model for generating cinematic clips from text and reference media. This page provides Seedance 2.5 Text-to-Video: turn one written prompt into a 4–30-second clip at 720p, with optional synchronized native audio generated with the picture. It accepts text only; use the other Seedance 2.5 pages when you need image, reference, or first-and-last-frame inputs.
| Seedance 2.5 advantage | What it means for you |
|---|---|
| Native audio in one pass | Generate synchronized speech, sound effects, and music alongside the video, so you can skip a separate dubbing and sound-design step. Turn it off when you only need silent video. |
| Up to 30-second clips | Direct a longer single shot with a clear beginning, development, and ending instead of stitching several short takes. |
| Stronger consistency and cleaner extensions | Seedance 2.5 targets steadier characters, wardrobe, and style, and curbs the blur that used to build up when a shot is extended. |
| Prompt-driven control | Follow negative and timestamp-style instructions more reliably for timed actions, framing, and multi-language scenes. |
9:16 Shorts, Reels, and Stories or 1:1 feed assets with a clear opening hook and platform-ready framing.The table below lists the controls exposed by the Seedance 2.5 Text-to-Video tool on this page.
| Parameter | Required | Type | Default | Range / Options | How to choose |
|---|---|---|---|---|---|
prompt* | Yes (*) | String | Example prompt | Chinese ~≤500 characters or English ~≤1000 words recommended | Give Seedance 2.5 a structured brief covering the subject, one main action, camera, setting and lighting, visual style, and any audio. |
aspect_ratio | No | String | 16:9 | 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive | Match the destination: 16:9 for general video and ads, 9:16 for vertical social, 1:1 for feeds, 21:9 for ultra-wide. adaptive lets the model pick the closest ratio. |
duration | No | Integer | 5 | 4–30 seconds, in 1-second steps | Use short clips for a single action; reserve longer durations for prompts that define a clear beginning, development, and ending. |
generate_audio | No | Boolean | true | true / false | Leave on for synchronized speech, effects, and music; turn off for silent video. |
\* Required field.
The following table summarizes the wider Seedance 2.5 family on RunComfy. The image, reference, and first/last-frame inputs are available on separate pages, not as controls on this Text-to-Video page.
| Seedance 2.5 page | Inputs on that page | Notes |
|---|---|---|
| Text-to-Video (this page) | prompt only, plus aspect_ratio, duration, generate_audio | Text-only; output at a fixed 720p (a 480p page is also available). |
| Image-to-Video | prompt + one image | Animates a still; output aspect ratio follows the input image. |
| Reference-to-Video | prompt + up to 9 images, 3 videos, 3 audios, plus aspect_ratio | References steer identity, motion, and mood. |
| First & Last Frame | prompt + first_frame + last_frame | Bridges two keyframes; output aspect ratio follows the frames. |
Every page runs 4–30-second clips and can generate optional native audio.
Seedance 2.5 Text-to-Video is billed per second of generated video at a fixed 720p:
| Duration | Price at $0.35/s |
|---|---|
| 5s | $1.75 |
| 10s | $3.50 |
| 15s | $5.25 |
| 30s | $10.50 |
For batches of multiple outputs, calculate the total as duration × $0.35 × output count.
Use this reusable Seedance 2.5 prompt structure:
[subject + defining details] + [one action over time] + [camera framing and movement] + [setting + lighting] + [visual treatment] + [dialogue / effects / ambience / music] + [ending frame or constraint]
Weak prompt
> A retro diner at night, cinematic, with rain.
Improved Seedance 2.5 prompt
> A retro American roadside diner glows at dusk, warm neon signs reading “DINER” and “OPEN” as light rain mists the empty parking lot and colorful reflections shimmer on the wet asphalt. Begin on a wide establishing shot, then a slow dolly-in as a classic car's headlights sweep past. Cinematic anamorphic look with soft lens flares; audio: gentle rain and distant thunder ambience, no dialogue.
The improved version gives Seedance 2.5 an identifiable subject, timed action, separate camera direction, lighting, finish, and sound sources.
Seamlessly lengthen shots with frame-consistent context control and audio blending for refined video creation.
Turn photos into expressive videos with synced voice motion.
LTX 2 retake video modifie the video by the prompt.
Easily add custom LoRA for unique styles and effects.
Seedance 2.5: Animate a still image into cinematic AI video
Generate video from multi-keyframe stills with optional audio
Yes. Seedance 2.5 Text-to-Video is available on RunComfy for browser testing and API integration.
Seedance 2.5 is ByteDance Seed's next-generation multimodal video model. Its broader family covers text, image, video, and audio inputs, while individual RunComfy pages expose different controls. This page provides the text-only Text-to-Video workflow.
Seedance 2.5 Text-to-Video turns prompts into short cinematic clips with optional native audio. It targets ad creative, film previsualization, and branded storytelling where prompt-driven control and synchronized audio matter.
Video is output at a fixed 720p on this page. Aspect ratio can be 16:9 (default), 9:16, 1:1, 4:3, 3:4, 21:9, or adaptive (the model picks the closest ratio). Duration is a whole number of seconds from 4 to 30, with a default of 5.
Yes. generate_audio is on by default, so the model can output synchronized speech, sound effects, and music, which supports lip-synced clips. Turn it off when you only need silent video. Audio quality still depends on prompt clarity and the scene you describe.
No. This page is the text-only Seedance 2.5 Text-to-Video workflow, and its API accepts only prompt, aspect_ratio, duration, and generate_audio. For source media, use the separate Seedance 2.5 Image-to-Video, Reference-to-Video, or First & Last Frame page.
No. RunComfy provides Seedance 2.5 through the browser and an HTTP API, so you do not need to host or scale the model yourself.
Prototype in the RunComfy model UI, then call the same model through the RunComfy API using identical Input fields (prompt, aspect_ratio, duration, generate_audio). Validate prompts in the UI first, then use your account API key and credits for automated jobs.
Generations are billed on the generated video duration at $0.35 per second at 720p. A 5-second clip costs about $1.75 and a 30-second clip about $10.50. See the Generation section on this page for live pricing.
RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.
